imdave@cbnewsc.ATT.COM (david.e.bodenstab) (07/24/89)
My upgraded 2.4.0L system will not recognize my second disk -- 2.3.0 worked fine. Has anyone run into this before? My system: NEC APC/IV (AT) disk 0: NEC 40M, 1 dos + 1 unix partition (root, usr, swap and tmp) disk 1: Seagate 4096 80M, 1 dos + 1 unix partition (usr/src and usr/local) When I booted 2.4, it wouldn't access disk 1. So, I backed everything off to diskettes, and ran `fdisk' to re-assign bad blocks and then `divvy 1' to make my file systems. Divvy worked fine. *But* when I tried to `labelit /dev/dsk/1s0' I got "cannot read superblock". It turned out I could not access disk 1 period. Then, I tried `divvy -u' which is supposed to update the "slice table" whatever that is. After doing this, the system was able to access disk 1 correctly. *But*, sometimes, after doing this there seems to be some sort of internal file system corruption for the root file system -- like "cannot stat .", "cpio: cannot execute", etc. After I reboot and run fsck, there apparently was no corruption on the disk. So, if I don't do a `divvy -u' I can't access my second disk, and if I do do a `divvy -u' I can access disk 1 but disk 0 is sometimes hosed. Could anyone shed some light on this? What am I doing wrong? I've even run `installit' from the 2.4.0 boot diskette and followed the instructions for installing on a two disk system (I didn't let divvy run mkfs on disk 0 though). Any help would be greatly appreciated. Thanks. Dave Bodenstab ...att!iwtmx!imdave 312 9798114
johnk@opel.UUCP (John Kennedy) (07/25/89)
In article <1939@cbnewsc.ATT.COM> imdave@cbnewsc.ATT.COM (david.e.bodenstab) writes: > >My upgraded 2.4.0L system will not recognize my second disk -- 2.3.0 worked >fine. Has anyone run into this before? > >My system: NEC APC/IV (AT) > disk 0: NEC 40M, 1 dos + 1 unix partition (root, usr, swap and tmp) > disk 1: Seagate 4096 80M, 1 dos + 1 unix partition (usr/src and > usr/local) > >When I booted 2.4, it wouldn't access disk 1. So, I backed everything off >to diskettes, and ran `fdisk' to re-assign bad blocks and then `divvy 1' to >make my file systems. Divvy worked fine. *But* when I tried to `labelit >/dev/dsk/1s0' I got "cannot read superblock". It turned out I could not >access disk 1 period. > ... >So, if I don't do a `divvy -u' I can't access my second disk, and if I do >do a `divvy -u' I can access disk 1 but disk 0 is sometimes hosed. This is so similar to what I've been going through it's frightening. Similar situation, where drive 1 appears to be causing failures on both disk 1 and disk 0. My old configuration: Disk 0 Disk 1 Seagate 4096 (80M) Seagate 251-1 (40M) This worked well. New configuration Disk 0 Disk 1 Seagate 4096 (80M) 1 Seagate 4096 (80M) #2 BY ADDING A SECOND 4096 I got those same brutal errors that you are describing. Either 4096 works find as drive 0 alone; either 4096 works along with the 251-1 as drive 1. I had concluded that either it was a fluky drive or uport 2.4 is having a problem with two 4096s in the system. Now your posting suggests that the problem could be related to having the 4096 as drive 1!!!! SO, anybody out there successfully running either TWO ST-4096s or running a 4096 as drive 1? I'd love to get this thing nailed down. BTW, my new 4096 has been returned to the vendor on the supposition it is faulty. John -- John Kennedy johnk@opel.UUCP Second Source, Inc. Annapolis, MD
rsj@wa4mei.UUCP (Randy Jarrett WA4MEI) (07/26/89)
In article <1939@cbnewsc.ATT.COM> imdave@cbnewsc.ATT.COM (david.e.bodenstab) writes:
++>
++>My upgraded 2.4.0L system will not recognize my second disk -- 2.3.0 worked
++>fine. Has anyone run into this before?
++>
++>My system: NEC APC/IV (AT)
++> disk 0: NEC 40M, 1 dos + 1 unix partition (root, usr, swap and tmp)
++> disk 1: Seagate 4096 80M, 1 dos + 1 unix partition (usr/src and
++> usr/local)
++>
++>When I booted 2.4, it wouldn't access disk 1. So, I backed everything off
++>to diskettes, and ran `fdisk' to re-assign bad blocks and then `divvy 1' to
++>make my file systems. Divvy worked fine. *But* when I tried to `labelit
++>/dev/dsk/1s0' I got "cannot read superblock". It turned out I could not
++>access disk 1 period.
++>
When you use divvy to setup the file systems on drive 1 don't use the
root entry. I had the same problems and they all cleared up when I used
the usr entry '/dev/dsk/1s2' for the system. I now have a larger drive and
am using all 4 of the file systems for drive 1 and not having any problems but
when it was the only one it wanted to be 1s2.
--
Randy Jarrett WA4MEI
UUCP ...!gatech!wa4mei!rsj | US SNAIL: P.O. Box 941217
PHONE +1 404 493 9017 | Atlanta, GA 30341-0217raw@ushiva.UUCP (Roland Wilcher) (07/30/89)
In article <2421@wa4mei.UUCP> rsj@wa4mei.UUCP (Randy Jarrett WA4MEI) writes: >I don't feel that the Seagate drives can hold up to the workout that unix >gives them. > >I switched over to Micropolis and CDC drives ( two different systems both >running V/AT ) and didn't have any problems out of either for over a year. I I have been running a 4096 for 2 Years with a news feed and have yet to have a single problem with the drive. Though a bit on the slow side, the drives seem fairly reliable in my experience. --- "If aint broke, don't fix it" raw@ushiva.ncoast.org -Bert lance uunet!hal.cwru.edu!ncoast!ushiva!raw 6203 Luther Ave. Cleve Oh. 44103 Roland A. Wilcher ------------------------------------------------------------------------------
mrm@Sceard.COM (M.R.Murphy) (07/30/89)
In article <2421@wa4mei.UUCP> rsj@wa4mei.UUCP (Randy Jarrett WA4MEI) writes: > >I tried to use some ST-4096 drives on my system with V/AT and out of about >6 drives that I tried I could only get one of them to work for any longer >than a couple of days and it only lasted for a couple of weeks before I >started getting seek errors on it also. > >I don't feel that the Seagate drives can hold up to the workout that unix >gives them. We've been using ST4096 for about 2 years (16 drives, 2/system) under SV/AT, WD1003-WA2. The only problem with a drive that we've had is when one system was thrown off the back of a truck. Then the drive made real funny screeching noises when it was turned on. No funny smell, just the noises... and lots of i/o errors. The drive is currently being rebuilt, just to see if that's a reasonable thing to do. Trying to use RLL controllers, the 4096 seemed too flakey to use. >[stuff deleted...] I >have now switched to two of the CDC Wren ESDI drives and am very happy with >the results. > Good to hear, thanks. --- Mike Murphy Sceard Systems, Inc. 544 South Pacific St. San Marcos, CA 92069 mrm@Sceard.COM {hp-sdd,nosc,ucsd,uunet}!sceard!mrm +1 619 471 0655
johnk@opel.UUCP (John Kennedy) (07/30/89)
Thanks to everyone acknowledging that they are successfully running ST-4096's under Sys V/AT. My question still remains. Is anyone using TWO of them, or using one as drive 1, rather than drive 0? My problem remains, after sending a drive back for exchange, that the drive runs fine by itself or with an ST-251, but two 4096's still will not coexist. Thanks, John -- John Kennedy johnk@opel.UUCP Second Source, Inc. Annapolis, MD
bill@twwells.com (T. William Wells) (07/31/89)
In article <2421@wa4mei.UUCP> rsj@wa4mei.UUCP (Randy Jarrett WA4MEI) writes:
: I don't feel that the Seagate drives can hold up to the workout that unix
: gives them.
On the other hand, I have an ST4096 on my Unix system, with a full
newsfeed on it. Other than a few blocks which intermittently failed
right after I installed it, I have had no problems.
The biggest problem I had was the *software* which would forget that
a block was bad if it ever managed to read it without error.
---
Bill { uunet | novavax | ankh | sunvice } !twwells!bill
bill@twwells.com